EXPLORERS STILL HOPE TO IMPROVE ON 3-D SEISMIC'S WEALTH OF DATA

Elwood O. Nestvold
P.H.H. Nelson Shell Internationale Petroleum Mij.
By The Hague

The occurrence of hydrocarbons is widespread, and reserves of oil have been discovered in some 90 countries.

However, these reserves are unevenly distributed; just four Middle Eastern countries contain more than half the world's proved oil reserves.

The fact remains that in spite of an intensive worldwide exploration effort, no major new petroleum provinces, such as the North Sea or Alaska, have been found in recent years.

Considerable amounts of oil have been discovered, but little of it in large fields of more than 500 million bbl. It seems clear, therefore, that new reserves will be increasingly in smaller less accessible accumulations. 1

BASIN MODELING

Maintaining reserves requires continual investment in exploration, much of which is naturally in areas where production is already established.

However, a significant effort also goes into searching for oil and gas in new areas. To promote this exploration, Shell Research has developed a two-dimensional basin modeling program.

The program simulates observed basin stratigraphy and predicts the distribution of lithofacies in areas with few wells and little seismic.

It operates by reconstructing the stratigraphy of a basin on a spatial grid, and in a sequence of small time steps from a prescribed set of initial conditions. These include a sea level curve, a thermal cooling curve, and bowl-shaped initial bathymetry. 2 3

GENERATION/MIGRATION

Focusing now on the pay, deterministic hydrocarbon-habitat modeling aims at predicting the quantity and distribution of hydrocarbons through geologic time. It can, therefore, play a major role in deciding where to drill.

In the sequence of processes leading to the eventual entrapment of hydrocarbons, secondary migration plays a key role. How, for instance, could one predict the outcome of drilling in a new basin before the first well is drilled?

To take one example, consider the results of a map-based modeling study of oil generation and migration off mid-Norway (Figs. 1, 2). 4 Charge was assumed to come only from the Kimmeridge clay equivalent, expelling a fight oil downward into the carrier bed, taken to be homogeneous and continuous over the study area.

Although simple, the model has proved useful for simulating oil migration and accumulation patterns.

The input required by the model's simulator consists of a structure-contour map (at the base of the topseal), rate maps of oil expulsion, and the properties of the carrier bed and oil for a geologically relevant time period.

The program output consists of calculated column heights below the seal. These may be plotted progressively in different color shades on the structure-contour map. 5

By providing the means to test alternative hypotheses, this simulation package promises to become a powerful prospect appraisal tool.

MODELING STRUCTURE

Having defined the existence of a basin likely to contain hydrocarbons, and having the means to simulate their generation and migration, it is essential to be able to define the three-dimensional structures in which they may be trapped.

This is where a thorough understanding of the dynamics of structural geology is a prerequisite. How do the complex trapping configurations (Figs. 3, 4) come into being?

In the seismic section (Fig. 3), clear-cut seismic amplitude anomalies and flat spots can be seen, revealing the presence of four stacked oil accumulations. These are located in the far distant corner of the isometric display (Fig. 4).

In displays from another 3-D seismic survey, the intricate distribution of oil in a reservoir deep within a collapsed crest anticline (Fig. 5) is revealed by the horizon amplitude map (Fig. 6). Here, the presence of oil in the reservoir increases the reflection strength of the seismic horizon.

These high amplitude areas are highlighted in red, providing the interpreter with a direct statement of the presence and detailed distribution of the oil in the subsurface, 6 7 8 9

Before dealing with the subject of 3-D seismic itself, it is useful to recognize that X-ray tomography and image processing of sandbox experiments can reveal the growth of complicated fault patterns in tectonic settings as diverse as Tertiary deltas (gravity tectonics), strike-slip terrains (wrench tectonics), and inversion terrains.10 The X-ray slices analyzed in the laboratory are strictly analogous to the 3-D seismic sections ultimately resulting from seismic surveys in the real world.

3-D VISUALIZATION

As an aid to visualizing complex structures in 3-D space-often difficult with only 2-D displays, recent experiments with interactive volume modeling on the computer have shown the potential for this new technology to allow geologists or geophysicists to check their interpretations in an economical and flexible way.

The input data set consists of re-sampled horizon and fault data from a completed 3-D seismic interpretation. The scattered data points look like stars in a night sky (Fig. 7).

At the flick of a switch, the data points can be suppressed, and the model can be fleshed out to give it more substance (Fig. 8).

With considerable ease and speed, the geologist can further flesh out the model by undoing the slicing option, used in the first figure to reveal the data points more clearly.

With equal ease and speed, the model may be rotated about various axes. The result of rotation around a horizontal axis, which reveals more clearly the faults affecting the upper surface of the model, and the color contouring of structural elevation, which generates a more vivid impression of the structural configuration, can be seen in Fig. 9.

Color contouring shows the throw variation along two intersecting fault planes (Fig. 10). Red indicates large throw, blue small throw.

The faults are of the collapsed crest type, the largest throw being in the center of the fault. Fault throw dies out upwards and downwards.

Faults can only be modeled with this software when they are treated as thin zones rather than discrete surfaces. Such fault planes are defined as volumes constrained by top and bottom surfaces.

Laterally the volumes are constrained by boundaries that have slight indentations because they follow the outermost extent of the scattered data points.

Two or more fault planes can be displayed in the same 3-D space, allowing intersecting relationships to be analyzed.

This kind of visualization capability greatly improves the understanding of geological and geophysical data, and inconsistencies in interpretations are quickly revealed.

Indeed, virtual reality may well become a prime quality control tool for 3-D seismic interpretation in the not too distant future.

3-D SEISMIC-IMPACT

These are indeed "boom times for 3-D seismic." 12

Shell Cos. can confidently lay claim to 25% of the 1,000 surveys reported to have been shot. 13 14 15

It is interesting to note that more than 50% of the cumulative 48,000 sq km acquired by Shell companies outside North America-in fact nearly 27,000 sq km-were acquired in 1989 and 1990. 16 17 18 19

This explosive growth has only been made possible by the development of imaginative new seismic acquisition, processing, and interpretation techniques. 20 21 22 23

The latest two boat acquisition technology involves two lines of air-gun pulses trailing astern of each boat. By using four streamers at a time, 16 subsurface lines can be acquired simultaneously and, with this technique, up to 50 sq km of full-fold 3-D coverage can be surveyed in 1 day.

IMPACT-OFFSHORE

3-D seismic has been found to have a significant impact on the ability of a company to explore its acreage effectively. 24 25 26 27

A case history from the North Sea show how dramatically our view of the subsurface has changed over the years, not because of the interpreters but because the 2-D grid and the 2-D data quality were inadequate.

Prospects in yellow show how the area looked in 1970 (Fig. 11). At that time no well had been drilled, and the 2-D seismic coverage was sparse. The biggest prospect was perceived to be in the west, so it was decided to drill a well there.

That well discovered gas shown on a 1974 map, but the well results and additional 2-D seismic showed the structure to be smaller than perceived in 1970. By 1974 it seemed that the largest prospects i ere in the center of the block, so a second well was planned on the largest remaining prospect.

Well No. 2 was successful; but again, the structure was radically reduced in size 9 years later (see cover of this issue). Three dry holes had been drilled in the west (Wells 3, 4, and 5), and there seemed to be few remaining prospects. There were then discussions about whether or not to relinquish the acreage, but instead it was decided to shoot a 3-D seismic survey over the eastern part of the block, considered to be the most prospective area at that time.

Two years later the situation had improved, since the new 3 D seismic survey in the east had revealed a number of prospects. Two of these were drilled as gas discoveries (Wells 6 and 7), and another 3-D survey was commissioned to cover the western part of the block.

By 1989, after covering the whole block with 3-D seismic, several additional prospects had been delineated. Six successful wells were drilled (8, 9, 10, and 11 in the east, and 12 and 13 in the west).

In 1990 two more wells were drilled. Well 14, in the north, was a dry hole, the first one since the 3-D seismic was acquired.

Possible reasons for the absence of hydrocarbons in this well are lack of gas charge due to late structuration or lack of closure due to uncertainties in the time-to-depth conversion.

Another well just outside the block in the southeast found gas but less than anticipated. This may also be due to limited gas charge, and a smaller closure in depth than in time.

The latest well, No. 15, has converted vet one more prospect (mapped on the basis of 3-D seismic) into a proven gas accumulation (last slide on cover), and still some undrilled prospects remain.

The success in this block and others resulted in the commissioning of large fairway 3-D surveys in the Dutch offshore. 28 29 30

IMPACT-ONSHORE

3-D seismic bears fruit onshore, as well as in the marine environment. 31 32 33 34 The next example, from onshore Holland, is illustrated by the difference between the 1974 interpretation based on a 2-D seismic grid of about 2 by 2 km and a more recent interpretation based on 3-D seismic data acquired in 1987 (Figs. 12, 13).

The well was drilled in 1990, resulting in the largest gas discovery in the northeast of the Netherlands since 1976.

The difference between the two maps is striking. The 2-D map shows only an insignificant closure at the location of the successful well.

Furthermore, no closure exists in the southwest of the 3-D map, where the 2-D map indicates a large prospect. In this case, the fault that forms the trap for this gas discovery is seen most clearly on the dip map of the objective horizon (Fig. 14).

The method whereby such a dip map is created is explained in the next section.

3-D INTERPRETATION

3-D seismic interpretation is now done exclusively on one or other of the commercially available interactive "workstations." 35 36

These sophisticated computers with their high resolution color screens enable the interpreter to analyze the entire data volume resulting from a 3-D seismic survey. Not only do the workstations permit a limited amount of 3-D visualization but more importantly they allow the precise measurement of all attributes of the seismic reflections that occupy the data volume. 37 38 39

After careful selection and/or combination of these attributes (such as reflection time, reflection strength, or reflection frequency content) maps may be made and displayed with the workstation that give direct insight to the geology of the subsurface, including in many instances the location of oil and gas (Fig. 6).

A time slice-a horizontal slice-through a 3-D data volume from a marine survey in the North Sea is shown (Fig. 15).40 The green loop has been picked as the top Jurassic reflection in this area.

The same green loop is shown in yellow (Fig. 16), together with two intersecting line tracks from the control, or seed grid. Notice that the intersecting lines are shaded in a variety of colors, indicating the change of reflection time of the top Jurassic horizon along these lines.

On the whole of the long east-west line, picked out in Fig. 16, the top Jurassic horizon has been highlighted in green (Fig. 17).

After manually interpreting a number of additional lines selected from the data set, the interpreter builds up the control or seed grid to the situation shown in Fig. 18.

This is where the full power of modern interpretation software is brought to bear, as the interpreter hands over the remainder of the task to the automatic volume picker. Fig. 19 shows an early stage in the process, which has been started in this case in the southeast corner of the data set.

The picking proceeds entirely automatically through the data set, and the end result is an horizon map of reflection time where every single data point has been picked and mapped (Fig. 20). In this case it means a data point every 25 m in each direction throughout an area of some 100 sq km.

With such a complete data file for every horizon interpreted in this way, it is a simple matter to calculate the attributes of dip, the inclination of the horizon, and azimuth, the direction in which it is inclined.41 42 The result of displaying the azimuth computations is shown (Fig. 21).

Shell Cos. have put considerable effort into developing this technique, which gives an immediate feel to the shape of the dome and the faults that cross it. The color code, something developed by Shell geophysicists, suggests the presence of low, red sunlight in the east and the cold blue light of the sky in the west.

In the corresponding dip display (Fig. 22), only the steepness of the dip is color coded, the steepest parts of the horizon being shown in red, and areas of very little dip shown in yellow. Green areas indicate intermediate values.

Some of the faults stand out particularly clearly in this display. They are picked out in red, naturally, since the horizon is interpreted to be very steep across such interruptions.

Also note that by combining the two images (Figs. 21, 22) the interpreter is provided with a total summary of the three dimensional geometry of the structure that he or she is looking at.

GEOLOGY REVEALED

Images from a 3-D land survey in New Zealand beautifully demonstrate the power of the 3-D seismic technique.

The azimuth map is derived from the time contour map of a key horizon in the depth range 11,000-13,000 ft (Fig. 23).

The azimuth map is shaded somewhat differently from the previous example. Here the warm red glow of the Sun is still in the east, and the western flank of the anticline is bathed in a steel gray light, more like that of the moon.

The north-facing plunge is brightly illuminated, and the combination of these shades reveals the subtle crestal faulting very clearly.

This image, generated only 10 days after loading the raw data into the workstation, revealed the presence of hitherto unknown crestal faults despite the fact that the field had already been in production for some 20 years.

More remarkable still is the amplitude map of this horizon-that is the variation of reflection strength (Fig. 24). Strong reflection energy is shown in red, and the weakest in black.

More or less in the center of this map a meandering black track can be clearly seen. There are good reasons to believe that this is the track of an early Tertiary river, which cut its way through the coal swamp preserved in the rest of the area as a thick coal seam.

This is believed to be a direct expression of paleogeography at 11,000-13,000 ft (3,350-3,960 m).

CONCLUSION

John Jennings, a Shell Group managing director, aptly put his finger on it when he identified 3-D seismic as, arguably, the single most important breakthrough in technology in recent years. 43

As far as our perception of the subsurface is concerned, it really is Eke replacing your clockwork gramophone with a compact disc player.

To be able to go on finding and exploiting what we might describe as conventional oil and gas-within the tight economic constraints likely to persist throughout the 1990s-all of the technology described here will certainly be required.

ACKNOWLEDGMENTS

The authors are grateful to many colleagues in Shell, both for continuing discussions on 3-D seismic and for assistance in providing materials for this article. We wish to thank J.H. Barwis, C. Bukovics, E.M. Daukoru, D.S. Davidson, D.J. Davies, Y.H. Eskes, D.J. Feenstra, J.G.A. Hillaert, W.T. Horsfield, J.K. Huyzinga, B.K. Levell, P.A.B. Marke, J.J. Nooteboom, J.M. Oosterbaan, M.P.A.M. Peters, H.J. Poelen, W.D. Poldermans, E.J.H. Rijks, C.K.J. Rutten, E. van Scherpenzeel, R.O. Speight, W. Voggenreiter, A.W. Wood, and P.R. Wood.

We are indebted to the following Shell and affiliated companies for permission to include information from their activities: Koninklijke/Shell Exploratie en Produktiie Laboratiorium (Shell Research BV, Maersk Olie og Gas AS, Nederlandse Aardolie Mij. By, Shell Todd Oil Services Ltd., and Shell Petroleum Development Co. of Nigeria Ltd.

Thanks are also due to Shell Internationale Petroleum Mij. By for permission to publish this article.

REFERENCES

The Shell Review, Shell International Petroleum Co. Ltd., Shell Centre, London, 1991, 60 pp.
Featherstone, P., Aigner, T., Brown, L., King, M., and Leu, W., Stratigraphic modeling of the Gippsland basin, APEA Journal, 1991, pp. 105-114.
Lawrence, D.T., Doyle, M., and Aigner, T., Stratigraphic simulation of sedimentary basins: concepts and calibration, AAPG Bull., Vol. 74, 1990, pp. 273-295.
Hermans, L., van Kuyk, A.D., Lehner, F.K., and Featherstone, P.S., Modeling secondary hydrocarbon migration in Haltenbanken, Norway: Koninklijke/Shell Exploratie en Produktie Laboratorium, Rijswijk, Netherlands, Publication 38 presented at NPF Conference, Stavanger, October 1989.
Lehner, F.K., Marsal, D., Hermans, L., and van Kuyk, A., A model of secondary hydrocarbon migration as a buoyancy-driven separate phase flow, Revue de l'Institut Francais du Petrole, Vol. 43, No. 2, 19887 pp. 155-164.
Bouvier, J.D., Kaars-Sijperstein, C.H., Kluesner, D.F., Onyejekwe, C.C., and van der Pal, R.C., Three-dimensional seismic interpretation and fault sealing investigations, Nun River field, Nigeria, AAPG Bull., Vol. 73, 1989, pp. 1,397-1,414.
Buchanan, R., Marke, P.A.B., and Ruijtenberg, P.A., Applications of 3-D seismic to detailed reservoir delineation, SPE Vol. 175, No. 5, 1988.
Buchanan, R., Ruijtenberg, P.A., and Marke, P.A.B., Applications of 3-D seismic to field development planning (abstract), AAPG Bull., Vol. 73, 1989, p. 338.
Gaarenstroom, L., The value of 3-D seismic in field development, SPE Paper 13049, 1989.
Koopman, A., Speksnijder, A., and Horsfield, W.T., Sandbox model studies of inversion tectonics, Tectonophysics, Vol. 137, 1987, pp. 379-388.
Wolfe, R.H., Jr., and Liu, C.N., Interactive visualization of 3-D seismic data: a volumetric method, IEEE Computer Graphics and Applications, Vol. 8, No. 4, 1988, pp. 24-30.
Nina Morgan in Petroleum Economist, July 1990.
Karlsson, T., Application of new 3-D technology, OGJ, Nov. 3, 1986, p. 46.
Ritchie, W., 1986, Role of 3-D seismic technique in improving oilfield economics, JPT, Vol. 38, 1986, pp. 777-786.
Ruijtenberg, P.A., Buchanan, R., and Markle, P.A.B., Three-dimensional data improve reservoir mapping, JPT, Vol. 42, 1990, pp. 22-25, 59-61.
Nelson, P.H.H., 3-D seismic: its growing use in exploration, Petroleum Exploration Society of Great Britain 25th Anniversary Conference, PETEX 1989 Abstracts, Paper 4.
Nestvold, E.O., The use of 3-D seismic in exploration, appraisal and field development, 12th World Petroleum Congress, Vol. 2, 1987, pp. 125-132.
Nestvold, E.O., Utilization de la Sismique 3-D dans l'Exploration, l'Appreciation et le Developpement des Gisements, Revue Tunisienne de I'Energie, Vol. 14, 1988, pp. 8-15.
Nestvold, E.O., The use of 3-D seismic in exploration and production, Shell International, presented at Energy Industries Council, London, Oil, Gas & Petrochemical seminar, Baghdad, Iraq, Oct. 14-16, 1989, 29 pp.
Davidson, D.S., and Bandell, A., A novel 3-D marine acquisition technique, 60th SEG annual meeting, San Francisco, 1990, expanded abstract, Vol. 1, pp. 863-866.
Davies, D.J., and Wood, P.R., New dimensions for oil and gas exploration, New Scientist, Vol. 1658, 1989, pp. 42-46.
Nooteboom, J.J., and Bukovics, C., Integrated 3-D land, shallow marine and deep channel seismic acquisition, 51st EAEG Conference, Berlin, 1989, abstracts, p. 19.
O'Connell, J.K., Kohli, M., and S.W. Amos, 1990, Bullwinkle: a unique 3-D experiment, 60th BEG annual meeting, San Francisco, 1990, expanded abstracts, Vol. 1, pp. 756-757.
Bouvier, J.D., Gevers, E.A.C., and Wigley, P.L., 3-D seismic interpretation and lateral prediction of Amposta Marino field (Spanish Mediterranean Sea), Geologie en Mijnbouw, Vol. 69, 1990, pp. 105-120.
Horvath, P.S., The effectiveness of offshore three-dimensional seismic surveys-case histories, Geophysics, Vol. 50, 1985, pp. 2,411-30.
Kluesner, D.F., Champion field: role of three-dimensional seismic in development geology of a complex giant oil field (abstract), AAPG Bull., Vol. 72, P. 207.
Schmidt, H.J.E,, Highlights of 3D surveys offshore Sarawak, presented at 7th SPE offshore South East Asia conference, Singapore, Feb. 2-5, 1988.
Nestvold, E.O., Eskes, Y.H., Nelson, P.H.H., and Rijks, E.J.H., 3-D seismic in exploration, 13th World Petroleum Congress, 1991, in press.
Nestvold, E.O., 3-D seismic: is the promise fulfilled?, 61st SEG annual meeting, Houston, 1991, expanded abstracts.
Oosterbaan, J.M., 3-D seismic as exploration tool, 60th SEC annual meeting, San Francisco, 1990, expanded abstracts, Vol. 1, pp. 160-161.
Blokzijl, J., IJssellmonde-Ridderkerk, The Netherlands: development of a mature oil field (abstract), AAPG Bull., Vol. 72, 1988, pp. 163-164.
Flint, S., Stewart, D.J., Hyde, T., Gevers, E.C.A., Dubrule, O.R.F., and van Riessen, E.D., Aspects of reservoir geology and production behavior of Sirikit oil field, Thailand: an integrated study using well and 3-D seismic data, AAPG Bull., Vol. 72, 1988, pp. 1,254-69.
Gardham, R., and W.W. Moeshart, 3-D seismic Ihsan/Jameel area (Oman), results and implications, SPE Paper 17995, 1989.
Jev, B.I., Kaars-Sijperstein, C.H., Peters, M.P.M.A., and Wilkie, J,T., The Akaso field, Nigeria: use of integrated 3-D seismic/fault-slicing/clay smearing on fault trapping and dynamic leakage (abstract), AAPG Bull., Vol. 75, 1991, pp. 602-603.
Berkhout, A.J., and Ridder, J., Workstations for integrated seismic interpretation, World Oil, Vol. 203, No. 5, 1986, pp. 33-38.
Gerhardstein, A.C., and Brown, A.R., Interactive interpretation of seismic data, Geophysics, Vol. 49, 1984, pp. 353-363.
Rijks, E.J.H., Attribute extraction: an important application in any detailed 3-D interpretation study (abstract), AAPG Bull., Vol. 74, P. 749.
Rijks, E.J.H. (presented by Jauffred, J.C.E.M.) Attribute ex- traction: an important application in any detailed 3-D interpretation study, 60th SEG annual meeting, San Francisco, 1990, expanded abstracts, Vol. 1, P. 774.
Rijks, E.J.H., and Jauffred, J.C.E.M., Attribute extraction: an important application in any detailed 3-D interpretation study, The Leading Edge, September 1991.
Brown, A.R., Graebner, R.J., and Dahm, C.G., Use of horizontal seismic sections to identify subtle traps, in AAPG Memoir 32, The deliberate search for the subtle trap, 1982, pp. 47-56.
Dalley, R.M., Gevers, E.C.A., Stampfli, G.M., Davies, D.J., Gastaldi, C.N., Ruijtenberg, P.R., and Vermeer, G.J.D., Dip and azimuth displays for 3-D seismic interpretation, First Break, Vol. 7, 1989, pp. 86-95.
Shell Internationale Petroleum Mij. By, Horizon processing techniques for recognition of structural geology on 3-D seismic, Research disclosure No. 29473, 1988.
Financial Times, Oct. 16, 1989.